uva-dare (digital academic repository) mass loss and ......van loon, j. t. (1999). mass loss and...

UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Mass loss and evolution of asymptotic giant branch stars in the MagellanicClouds

van Loon, J.T.

Publication date1999

Link to publication

Citation for published version (APA):van Loon, J. T. (1999). Mass loss and evolution of asymptotic giant branch stars in theMagellanic Clouds.

General rightsIt is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s)and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an opencontent license (like Creative Commons).

Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, pleaselet the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the materialinaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letterto: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. Youwill be contacted as soon as possible.

Download date:24 May 2021

https://dare.uva.nl/personal/pure/en/publications/mass-loss-and-evolution-of-asymptotic-giant-branch-stars-in-the-magellanic-clouds(63f5b4dc-1d00-4c84-9e9e-ad3ee63bf819).html

Chapterr 7

ISOO observations of obscured Asymptoticc Giant Branch stars in the Largee Magellanic Cloud

NormanNorman R. Trams, Jacco Th. van Loon, L.B.F.M. Waters, Albert A. Zijlstra, Cecile Loup, PatriciaPatricia A. Whitelock, M.A.T. Groenewegen, Joris A.D.L. Blommaert, Ralf Siebenmorgen, A. Heske,Heske, Michael W. Feast, 1999, A&A Main Journal in press

Wee present ISO photometric and spectroscopic observations of a sample of 57 bright Asymptotic Giantt Branch stars and red supergiants in the Large Magellanic Cloud, selected on the basis of IRASS colours indicative of high mass-loss rates. PHOT-P and PHOT-C photometry at 12, 25 andd 60 ^m and CAM photometry at 12 /xm are used in combination with quasi-simultaneous ground-basedd near-IR photometry to construct colour-colour diagrams for all stars in our sam-ple.. PHOT-S and CAM-CVF spectra in the 3 to 14 /im region are presented for 23 stars. From thee colour-colour diagrams and the spectra, we establish the chemical types of the dust around 499 stars in this sample. Many stars have carbon-rich dust. The most luminous carbon star in thee Magellanic Clouds has also a (minor) oxygen-rich component. OH/IR stars have silicate absorptionn with emission wings. The unique dataset presented here allows a detailed study of aa representative sample of thermal-pulsing AGB stars with well-determined luminosities.

7.11 Introductio n

Onee of the least expected achievements of the Infra-Red Astronomical Satellite (IRAS; Neuge-bauerr et al. 1984) was the detection of a large number of mid-IR point sources in the Large Magellanicc Cloud (LMC) just above its limits of sensitivity (IRAS Point Source Catalogue; Schweringg k Israel 1990). Many of these are candidates for intermediate-mass stars at the tip off the Asymptotic Giant Branch (AGB). Their lives drawing to a close, these stars are shedding theirr stellar mantles at rates of up to 10~4 M 0 yr-1. Their dusty circumstellar envelopes (CSEs) obscuree the optical light from the star and become very bright IR objects. The details of the evolutionn and mass loss of AGB stars are poorly understood. The study of galactic samples off AGB stars is severely hampered by the difficulty to determine accurate distances to stars in thee Milky Way. The distance to the LMC, however, is well known and hence luminosities and mass-losss rates of AGB stars in the LMC may be determined with a high degree of accuracy.

81 1

82 2 ChapterChapter 7

Tablee 7J IRAS detected stars observed with ISO: names (LI stands for LI-LM C (Schwering & Israel 1990), TRM is from Reid et al. (1990), HV iss from Payne-Gaposchkin (1971), SP is from Sanduleak & Philip (1970) and WOH is from Westerlund et al. (1981); sources will be referenced hereafterr by their bold-faced names), ISO pointing coordinates (J2000), and references: 1: Hodge & Wright (1969); 2: Eggen (1971); 3: Wright && Hodge (1971); 4: Dachs (1972); 5: Sandage & Tammann(1974); 6: Glass (1979); 7: Humphreys (1979); 8: Blanco et al. (1980); 9: Feast el al.(1980);; 10: Bessell & Wood (1983); 11: Wood et al. (1983); 12: Rebeirot et al. (1983); 13: Prevot et al (1985); 14: Elias et al. (1985); 15: Woodett al. (1985); 16: Elias et al (1986); 17: Woodet al. (1986); 18: Reid et al. (1988); 19: Reid (1989); 20: Hughes (1989); 21: Hughes & Woodd (1990); 22: Reid et al. (1990); 23: Hughes et al. (1991); 24: Wood et al. (1992); 25: Roche et al. (1993); 26: Groenewegen et al (1995); 27:: Zijlstra et al. (1996); 28: Ritossaet al. (1996); 29: van Loon et al. (1996); 30: van Loon et al. (1997); 31: Loup et al. (1997); 32: Oestreicher (1997);; 33: van Loonet al. (1998a); 34: Groenewegen & Blommaert (1998); 35: van Loon el al. (1998b); 36: van Loon et al. (1999)

LI I

1825 5 1844 4 4 4 57 7 60 0 77 7 92 2 121 1 141 1 153 3 159 9 181 1 198 8 203 3 297 7 310 0 383 3 570 0 571 1 578 8 612 2 1880 0 663 3 793 3

----1157 7 1130 0 1145 5 1153 3 1164 4 1177 7 1238 8 1281 1 1286 6 1345 5 1382 2 1506 6 1756 6 1790 0 1795 5

IRAS S

04286-6937 7 04374-6831 1 04407-7000 0 04496-6958 8 04498-6842 2 04509-6922 2 04516-6902 2 04530-6916 6 04539-6821 1 04544-6849 9 04545-7000 0 04553-6825 5 04557-6753 3 04559-6931 1 05003-6712 2 05009-6616 6 05042-6720 0 05112-6755 5 05113-6739 9

--05128-6728 8 05128-6455 5 05148-6730 0 05190-6748 8

----05295-7121 1 05289-6617 7

--05294-7104 4 05298-6957 7 05300-6651 1 05316-6604 4 05327-6757 7 05329-6708 8 05348-7024 4 05360-6648 8 05402-6956 6 05506-7053 3 05558-7000 0 05568-6753 3

TRM M

--------------------------------48 8 4 4 24 4 72 2 43 3

--36 6 20 0 88 8 45 5

--99 9

------79 9 101 1 5 5 60 0

--77 7

--------

HV V

--------------------------12501 1

----888 8

------2360 0

--916 6

----------5870 0

--------996 6

--------------

RAA (2000)

044 28 30.3 044 37 22.8 044 40 28.4 044 49 18.6 044 49 41.4 044 50 40.2 044 51 28.4 044 52 45.3 044 53 46.3 044 54 14.4 044 54 09.8 044 55 10.1 044 55 38.9 044 55 41.6 055 00 18.9 055 01 03.8 055 04 14.3 055 11 10.1 055 11 13.7 055 11 41.2 055 12 46.4 055 13 04.6 055 14 49.9 055 18 56.7 055 20 20.9 055 28 16.3 055 28 40.8 055 29 02.6 055 29 03.5 055 28 47.8 055 29 24.5 055 30 04.2 055 31 45.9 055 32 36.0 055 32 52.5 055 34 16.1 055 36 03.3 055 39 44.6 055 50 09.1 055 55 20.8 055 56 38.7

Decll (2000) Otherr names IRASIRAS detected stars

- 699 3049 - 688 25 03 -699 55 13 -699 53 14 -688 37 50 -699 17 33 - 688 57 53 - 699 1153 - 688 16 12 - 688 44 13 - 699 56 00 - 688 20 35 - 677 49 10 - 699 26 25 - 677 08 02 - 666 1240 - 677 16 17 - 677 52 17 - 677 36 35 -66511 12 - 677 19 37 - 644 51 40 - 677 27 19 - 677 45 06 - 666 36 00 -677 20 55 -711 19 13 - 666 1531 -699 06 47 -71022 29 -699 55 14 -666 49 23 -666 03 51 -677 55 08 -677 06 25 -700 22 53 - 666 4647 -699 55 18 - 700 53 12 -70000 05 -677 53 39

------------------SP777 30-6.

--WOHG64 4

--SP777 31-20,

----SP777 29-33.

------SP777 37-24,

--SP777 37-35,

----------SP777 47-17.

------

WOHH SG66

WOHH SG097

WOHH SGI40

WOHH SGI93

WOHH SG204

WOHH SG331

WOHH SG374 SP777 46-59,

--------------

WOHH SG388

References s

27.31,33 3 27,31,33 3 27,31,33 3 27,31,33,34,36 6 27,31,33,36 6 24.27.28,33,36 6 24,27,33 3 24.27.28 8 27.31,33 3 12,13,20,21,27,31 1 24.27 7 16,17,19,24,25,27.29.31.33.35,36 6 27,31,33 3 11,12.13,20,22,27.32,33 3 27.30.31,33,36 6 27.31,33.36 6 5.6,7,11.14.18.22.31,32 2 22,27.31.33.36 6 22.27.31.33 3 22,27.31.33 3 3,5,6,7,11,14,18,22,31 1 27,31,33.36 6 1,2,4,6,7,11,18,22,31,32 2 22.27,31.33 3 22.27,31.33,34,36 6 22.27.31.33 3 27,31 1 22.27.31 1 9,11,20,31 1 24,27.33 1 24,27,31.36 6 22,27.31,36 6 22,27.31 1 7,11,17,18,22.31 1 17,22.24,26,27,31.33.36 6 27,31,33 3 22.27,31,33 3 24,27 7 27.31.33 3 27,31 1 27,31 1

Earlyy explorations of the IRAS data in combination with ground-based near-IR observations resultedd in the first identifications of mid-IR sources in the LMC with obscured AGB stars (Reidd et al. 1990; Wood et al. 1992). We have successfully increased the sample of known AGB counterpartss of IRAS sources in the LMC from a dozen to more than 50 stars (Loup et al. 1997;; Zijlstra et al. 1996: papers I k II ; van Loon et al. 1997, 1998a: Chapters 2 k 3). We attemptedd to classify their photospheres and CSEs as oxygen- or carbon-dominated, but for the majorityy of the stars this could not be done conclusively There remained therefore considerable

ISOISO observations of AGB stars in IMC 83 3

Tfcblee 7.2 The list of program stars without IRAS counterpart. The references are as in Table 7.1. SHV is from Hughes (1989), BMB is from Blancoo et al. (1980), WBP is from Wood et al. (1985) and GRV is from Reid et al. (1988).

HV V RA(2000) ) Decll (2000) Otherr names References s rwn-IRASrwn-IRAS sources classified as C stars

--_ _ 2379 9 _ _ _ _ _ _ --

2446 6

_ _ _ _ _ _ --12070 0

044 53 59.7 055 02 28.7 055 14 46.3 055 2046.8 055 25 30.6 055 26 17.4 055 35 11.4

non-IRAS non-IRAS 055 20 01.5 055 21 33.1 055 2140.5 055 24 31.3 055 30 00.3 055 52 27.8

-677 45 47 -699 20 10 -677 55 47 -699 0125 -700 09 13 -699 08 07 -70222 46

SHV0454030-675031 1 SHV0502469-692418 8

--SHV0521050-690415,, BCB-R046 SHV0526001-701142 2 WBP14 4 SHV0535442-702433 3

sourcessources classified as Mor S stars -677 34 43 -70099 56 -700 22 31 -69433 25 -70200 06 -699 14 12

WOHH G274, GRV0520-6737 SHV0522023-701242 2 SHV0522118-702517 7 SHV0524565-694559 9 SHV0530323-702216 6 WOHSG515 5

20,21 1 20,21,23 3 3,10,11,20 0 8,20,21,23 3 20,21 1 15 5 20,21 1

11,18 8 20,21 1 20,21 1 20,21 1 20,21,23 3 9,11 1

non-IRASnon-IRAS sources without spectral classification

------

055 00 11.2 055 00 13.5 055 19 41.8

- 688 12 48 -688 24 56 -666 57 50

SHV0500193-681706 6 SHV0500233-682914 4 GRV0519-6700 0

20,21 1 20,21 1 18 8

uncertaintyy about the luminosity distribution of the obscured carbon stars. This information is importantt for testing current understanding of the evolution of AGB stars, including dredge-up off carbon and nuclear burning at the bottom of the convective mantle (Hot Bottom Burning, HBB). .

577 obscured AGB stars and a few red supergiants (RSGs) in the LMC were selected for Guaranteedd Time and follow-up Open Time observations with the Infrared Space Observatory (ISO;; Kessler et al. 1996). The goals were to obtain photometry at 12, 25 and 60 /zm and to spectroscopicallyy determine the chemical types of the CSEs. The photometry, which covers thee entire spectral energy distributions (SEDs), can be modelled and used to derive accurate luminositiess and mass-loss rates. In this paper we present the ISO data and classify sources as oxygen-- or carbon-rich.

7.22 Source selection

Thee sources observed with ISO were selected from the lists presented in Paper I, where all IRASS candidate AGB stars in the MCs are listed. We selected 30 infrared AGB stars or RSGs withoutt optical counterparts from their Table 2. These objects should represent the brightest, mostt obscured AGB stars. Four objects from this table were excluded because of their red IRASS colours (S25/Si2£2.5): LI-LMC528, 861, 1137 and 1341. We also selected 8 sources fromm the optically known M and C stars with IRAS counterparts in Table 1 of Paper I. These includee well known Harvard Variables as well as the optically thick source IRAS04553-6825 (LI-LMC181,, WOH G64; Elias et al. 1986; Wood et al. 1986). Two unidentified IRAS sources fromm Table 4 of Paper I have been included in the present sample. LI-LMC203 is near an Ml.5 starr (HV12501), but there is also an A3 lab supergiant (Sk-69-39a) close to the IRAS position. Forr LI-LMC1795 we found a bright R-band counterpart (Paper II) . Finally one source from Tablee 7 of Paper I was included (LI-LMC1130). Although listed in Paper I as a foreground star,, it was included here in an attempt to establish whether this is true. For these last three starss the higher spatial resolution ISO observations at 12 yum allow a better identification of


thee source with one of the possible counterparts found near the IRAS position. The 41 IRAS sourcess included in this study are listed in Table 7.1, with the most common names for these objects,, their coordinates (J2000) and some references. The coordinates for the pointings of thee ISO observations were taken from the SIMBAD astronomical database in 1994.

Thee selection of IRAS detected AGB stars gives a sample that is severely biased towards veryy luminous stars (including supergiants). We therefore also included 16 non-IRAS stars. Thesee were mostly taken from Wood et al. (1983, 1985), Reid et al. (1990) and Hughes (1989). Sevenn of these objects are classified as C stars from optical spectra or near-IR colours. Six objectss are classified as M or S stars and for three objects no classification is available. This groupp of non-IRAS sources also includes the RCB-like variable HV2379 (Bessell k Wood 1983). Thesee sources are listed in Table 7.2.

7.33 IRAS data

Heree the IRAS data are discussed, for later comparison with the ISO photometry. Data at 12, 25,, 60 and 100 /im was retrieved from the IRAS data base server in Groningen1 (Assendorp et al.. 1995). The Groningen Gipsy data analysis software was used to measure the flux density fromm a trace through the position of the star (Gipsy command SCANAID). For the 60 and 100 ^mm data, 2 x2 square degree maps were created with 0.5' pixels to find point sources coincident withh the positions of the stars. The 12 and 25 yum flux densities have a 1-CT error of a few per cent,, with a minimum error of ~ 0.01 Jy. The 60 and 100 /im flux densities are much less certain,, and it is also more difficul t to assess reliable error estimates: 10% would be a typical error.. The faintest 60 //m sources that IRAS detected were assigned F60 = 0.1 Jy. Only one sourcee was well detected at 100 /im. The flux densities are listed in Table 7.3. When it is not certainn that the measured flux density is physically related to the star of interest it is marked withh a colon.

Al ll of our sources that are in the IRAS-PSC, plus HV5870 (=LI-LMC1145) and TRM72 (=LI-LMC578)) that are in Schwering &; Israel (1990), were recovered with good flux den-sityy determinations at 12 /im. Reliable 12 /im flux densities could also be determined for IRAS05128-64555 and 05289-6617, below their upper limit s as listed in the PSC. Neither inn the PSC, nor in Schwering & Israel (1990), are HV12070, HV2379, HV2446, TRM45 and TRM888 secure detections. Detection is not certain for WBP14 and the SHV sources for which fluxx density estimates are listed. The 12 /im flux densities of the GRV source and four SHV sourcess are upper limits. IRAS05506-7053 looks extended or multiple.

Att 25 /im detections seem a littl e more reliable than at 12 /tin, at a given flux density. Ratherr surprisingly, the detection limi t at 25 /im is at least as faint as at 12 /im; sources with FF2255 ~ 0.02 Jy could be found (see also Reid et al. 1990). This is, however, only possible because thee positions of the stars are known. For SP77 3 0 -6 and all eight (other) IRAS sources the PSCC lists only upper limit s of F25 < 0.25 Jy. SHV0502469-692418 and WrBP14 were the only sourcess that were (tentatively) detected at 12 //m but not at 25 /im. Their flux densities are probablyy below the limi t of detection, F2s < 0.01 Jy, if their colours are rather blue.

Att 60 /xm, IRAS05298-6957 is a bright, small but extended source about 10' in diameter,

lrThee IRAS data base server of the Space Research Organisation of the Netherlands (SRON) and the Dutch Expertisee Centre for Astronomical Data Processing is funded by the Netherlands Organisation for Scientific Researchh (NWO). The IRAS data base server project was also partly funded through the Air Force Office of Scientificc Research, grants AFOSR 86-0140 and AFOSR 89-0320.

ISOISO observations of AGB stars in LMC 85

Tablee 73 Revised IRAS 12, 25,60 and 100 j*m photometry (in Jy), accompanied by a colon if questionable.

Star r GRV0519-6700 0 HV12070 0 HV12501 1 HV2360 0 HV2379 9 HV2446 6 HV5870 0 HV888 8 HV916 6 HV996 6 IRAS04286-6937 7 IRAS04374-6831 1 IRAS04407-7000 0 IRAS04496-6958 8 IRAS04498-6842 2 IRAS04509-6922 2 IRAS04516-6902 2 IRAS04530-6916 6 IRAS04539-6821 1 IRAS04545-7000 0 IRAS04557-6753 3 IRAS05003-6712 2 1RAS05009-6616 6 1RAS05112-6755 5 IRAS05113-6739 9 IRAS05128-6455 5 IRAS05190-6748 8 IRAS05289-6617 7 IRAS05294-7104 4 IRAS05295-712I I IRAS05298-6957 7 IRAS05300-6651 1 IRAS05329-6708 8 IRAS05348-7024 4 IRAS05360-6648 8 IRAS05402-6956 6 IRAS05506-7053 3 IRAS05558-7000 0 IRAS05568-6753 3 SHV0454030-675031 1 SHV0500193-681706 6 SHV0500233-6829I4 4 SHV0502469-692418 8 SHV0521050-690415 5 SHV0522023-701242 2 SHV0522118-702517 7 SHV0524565-694559 9 SHV0526001-701142 2 SHV0530323-702216 6 SHV0535442-702433 3 SP777 30 -6 TRM45 5 TRM72 2 TRM88 8 WBP14 4 WOHG64 4 WOHH SG374

F\2 F\2 <0.06 6

0.06 6 0.23 3 0.38 8 0.05 5 0.05 5 0.30 0 0.58 8 0.44 4 0.71 1 0.28 8 0.24 4 0.76 6 0.31 1 1.33 3 0.89 9 0.86 6 2.07 7 0.22 2 0.46 6 0.24 4 0.43 3 0.28 8 0.46 6 0.25 5 0.23 3 0.39 9 0.16 6 0.69 9 0.23 3 0.85 5 0.28 8 0.74 4 0.58 8 0.21 1 0.71 1 0.28 8 0.85 5 0.35 5

<0.03 3 0.11 1 0.10 0 0.02 2 0.06 6

<0.10 0 0.06 6

<0.14 4 0.07 7

<0.04 4 0.01 1 0.26 6 0.07 7 0.22 2 0.17 7 0.01 1 8.45 5 0.37 7

^ 2 5 5

<0.02 2 0.03 3 0.06 6 0.35 5 0.02 2 0.02 2 0.17 7 0.29 9 0.23 3 0.53 3 0.20 0 0.12 2 0.76 6 0.19 9 0.89 9 0.86 6 0.55 5 5.09 9 0.12 2 0.83 3 0.14 4 0.33 3 0.14 4 0.33 3 0.14 4 0.24 4 0.25 5 0.39 9 0.56 6 0.08 8 1.38 8 0.17 7 1.23 3 0.16 6 0.09 9 1.02 2 0.16 6 0.80 0 0.43 3

<0.03 3 0.07 7 0.03 3

<0.03 3 0.02: :

<0.04 4 0.05 5

<0.07 7 0.01: :

<0.04 4 0.07: : 0.13 3 0.07 7 0.06 6 0.04 4

<0.03 3 13.53 3 0.38 8

FeoFeo F100 <0.1 1

0.1: : <1.0 0

0.4: : <0.8 8 <0.2 2 <5.0 0 <4.0 0 <2.0 0 <0.5 5 <0.1 1

0.1: : 0.1 1 0.1 1

<0.2 2 <2.0 0

0.4: : 22.00 28.0 0.1: :

<0.5 5 <0.3 3

0.1: : <0.4 4

0.7: : 0.1: : 0.1 1 0.1: : 0.3 3

<3.0 0 <0.3 3 <3.0 0

0.1: : 0.2: :

< l .0 0 0.3: :

<2.0 0 <0.2 2

0.2: : 0.2 2

<0.2 2 <0.3 3 <1.5 5 <0.1 1 <0.7 7

0.4: : <2.2 2 <1.0 0

0.1: : 0.4 4

<1.0 0 0.1: :

<2.0 0 <0.3 3 <0.7 7 <4.0 0

2.2 2 0.2: :

8C C ChapterChapter 7

22 -

11 1.5

- 0 . 5 5

~i~i—i—i—I—I—i—i—I—i—i—i—I—i—i—i—I—i—r —i—i—I—I—i—i—I—i—i—i—I—i—i—i—I—i—r II ' ' ' I

oo * 2 . o * o

JJ i i i I i i i I i i i L

II I I I I —

= 12 yu.m oo = 25 /zm

1.5 5

0.5 5

0.22 0.4 0.6 0.8 10 0

revised d (Jy) )

Figuree 7.1: A comparison of the estimates of IRAS flux densities given here and values from thee IRAS Point Source Catalogue and Schwering & Israel (1990), at 12 (solid symbols) and 25 fimm (open symbols).

withh F60 ~ 2 Jy. No point source could be distinguished on top of this emission, that is probablyy associated with the small cluster of which IRAS05298-6957 is a member (Chapter 3).. Flux densities are listed for two dozen sources, but it is not sure how many among these aree real detections and how many are spurious. The only 60 fim detections in the PSC are IRAS04516-69022 9 Jy), 04530-6916 (20.51 1.85 Jy) and 05112-6755 (0.91 1 Jy),, all consistent with our estimates. More stringent upper limit s are put on the 60 fim flux densitiess of the other sources.

Att 100 fim sources may be detected as faint as a few Jy. The only detection, however, is thee brightest far-IR source in our sample, IRAS04530-6916, which we measured at F10o = 28 Jy.. This is consistent with the PSC upper limi t of 46.17 Jy.

Thee new flux density estimates can be compared with the literature values from the PSC orr Schwering & Israel (1990) (Fig. 7.1). On average, the new flux densities are only a few per centt fainter than the values from the literature. Flux densities F2 5^0.2 Jy may have been over-estimatedd in the past. HV12501 with Fr e v/F l i t = 0.56 and 0.32 at 12 and 25 fim, respectively, andd IRAS05506-7053 with Frev/Fm = 0.67 and 0.42 at 12 and 25 fim, respectively, are the mostt extreme examples of this. Schwering & Israel (1990) over-estimated the 25 ^m flux densityy of TRM72 (FTev/F\it = 0.55), but under-estimated the 12 fim flux density of HV5870 (Frev/Flitt = 2.00). The other two flux densities which are obviously under-estimated are for SP777 3 0 -6 at 12 ^m (Fr e v/F l i t = 1.53) and IRAS04286-6937 at 25 /jra (Fr e v/F,i t = 1.67).

IRASS counterparts not listed in IRAS-based catalogues may still be found in the original IRASS data. This is because manual extraction and measuring of the data is a more sophisticated techniquee than the automatic techniques that created the existing catalogues. In particular, manuall flux-density determination enables the background flux levels to be estimated and subtractedd better, yielding more reliable photometry. The only sources from our ISO sample thatt could not be detected in the IRAS data at either 12, 25 or 60 fim are GRV0519-6700. SHV0454030-6750311 and SHV0524565-694559.

ISOISO observations of AGB stars in LMC 87 7

7.44 ISO observations

Thee programme stars were observed with ISO at 12, 25 and 60 /im (chopped measurements) andd with PHOT-S as part of a Guaranteed Time programme under proposals NTMCAGB1 andd NTMCAGB2, and at 60 /im with mapping mode and CAM-CVF as part of an Open Time programmee under proposal LMCSPECT.

Thee 12 /im photometry was mostly obtained using the ISOCAM instrument (Cesarsky et al.. 1996) in staring mode with a 3" pixel field of view in beam LW-s and using the LW10 filter (~~ IRAS 12 /im). We did 25 exposures of each 2.1 s duration, after a number of read-outs too stabilise the detector (ranging from 10 to 34 frames depending on the expected source flux density).. The gain was 2 in most cases, but 1 in the case of sources that were expected to be relativelyy bright: HV12501 and 996, IRAS04496-6958, 04545-7000, 05003-6712, 05112-6755, 05348-70244 and 05568-6753, and WOH SG374. For most stars this resulted in a clear detection withh S/N ratios of 10 to 100. In total 44 sources were observed with ISOCAM at 12 /im.

Forr sources that were expected to be stronger than 0.5 Jy and which would therefore sat-uratee the ISOCAM detectors with the LW10 filter, the 12 /im photometry was obtained with thee ISOPHOT instrument (Lemke et al. 1996) using the 11.5 filter (~ IRAS 12 /im). These observationss were done using triangular chopping with a chopper throw of 90". The aperture usedd for the observations was 52" in diameter. Integration times were 32 s on-source (and the samee off-source), except for IRAS05294-7104 that we integrated 64 s. A total of 13 sources weree observed in this mode. For 53 sources we obtained PHOT-P photometry at 25 /im us-ingg the 25 filter (~ IRAS 25 /im), triangular chopping with a chopper throw of 90", and an aperturee of 52". Integration times ranged from 32 to 256 s, depending on the expected flux density.. In our Guaranteed Time programme we finally observed 40 objects with ISOPHOT att 60 /im using the PHOT-C100 camera and filter 60 (~ IRAS 60 /im) and triangular chop-pingg with a 150" chopper throw. Integration times ranged from 32 to 128 s, depending on the expectedd flux density. Unfortunately due to the reduced in-orbit sensitivity of the instrument andd the problems with the calibration of the chopped measurements (especially for PHOT-C), wee discovered after most observations had already been carried out that this was not the best observingg strategy for the 60 //m photometry. Therefore, 7 objects were observed again in the Openn Time using PHOT-C100 and filter 60 in raster mapping mode, with 3 x 3 rasters and 45"" raster steps in X and Y directions (spacecraft coordinates). The integration time per raster pointt was 128 seconds.

Inn order to establish the carbon- or oxygen-rich nature of some of the programme stars wee also obtained IR spectra for a number of them. In the Guaranteed Time 15 objects were observedd using PHOT-S in staring mode, with integration times of 256 or 512 s (1024 s for HV2379)) depending on the expected flux densities. The advantage of this instrument is that itss spectral coverage is rather large (2 to 12 /im) at a reasonable resolution (~ 90). The sensitivityy of the PHOT-S instrument, however, limits the detectability to sources with 12 /im fluxflux densities above ~ 0.3 Jy. Furthermore, using staring observations the background cannot easilyy be determined. In this spectral region the diffuse emission is dominated by the zodiacal emission,, which, according to IRAS measurements, amounts to about ~ 0.1 Jy in the PHOT-S aperturee at 10 /xm.

Consideringg this, we decided to obtain CAM-CVF spectra for 12 objects with a pixel field-of-vieww of 6" in beam LW-1. We did 25 exposures of 2.1 s each at gain 2, after 50 read-outs too stabilise the detector. The unprecedented sensitivity of the ISOCAM instrument allows the observerr to obtain spectra even for sources as faint as 100 mJy at 12 /xm. Because of the long


durationn of a CVF observation, the spectral coverage chosen was only 7 to 9.2 //m (with step 4)) in LW-CVF1 and 9 to 14.1 fim (with step 2) in LW-CVF2, at a spectral resolution of ~ 40. AA big advantage of the CAM-CVF is that the spectra are obtained using an imaging technique. Therefore,, a background spectrum was obtained simultaneously. These background spectra can bee used to correct the PHOT-S spectra. We also obtained observations of 3 objects for which PHOT-SS spectra had already been taken, in order to cross-check the results from the different instruments. .

7.4,11 Near-IR photometry

Near-IRR photometry was determined for each star at the time of the ISO observation, by inter-polatingg near-IR lightcurves from our monitoring campaign at the South African Astronomical Observatoryy (SAAO) at Sutherland, South Africa (Whitelock et al., in preparation). Nearly alwayss the lightcurve was sampled close in time to the ISO observation, but occasionally some extrapolationn was necessary. The quoted uncertainties include an estimate, for each star, of thee error introduced by the inter/extrapolation. For TRM45 and for the H-band magnitude off IRAS05360-6648 we have made use of the near-IR lightcurves and photometry presented byy Wood (1998), after transformation to the SAAO system using Carter (1990). The near-IR photometryy is listed in Table 7.4, along with the ISO photometry and the Julian Dates of the ISOO spectroscopy.

Noo near-IR counterparts could be identified with IRAS05568-6753 and 05289-6617. Two starss with near-IR colours much like those of unobscured M-type stars were monitored in the near-IR,, but they show no variability.

7.55 ISO results and comparison with IRAS photometry

Thee data were reduced using the PHOT and CAM Interactive Analysis software packages: PIA (Gabriell et al. 1997) version V7.1.2(e) and CIA (Ott et al. 1996) version V3.0, respectively. Forr a general description of the data and reduction methods see the ISOPHOT Data Users Manuall (Laureijs et al. 1998), and the ISOCAM Observer's Manual (1994) and ISOCAM Data Userss Manual (Siebenmorgen et al. 1998). Details of the steps undertaken in reducing so-called Editedd Raw Data (ERD) products to the finally derived flux densities and spectra can be found inn Appendices C (photometry) and D (spectroscopy). The resulting ISO photometry is listed inn Table 7.4, and the ISO spectra are presented in Figs. 7.4 & 7.5.

Thee flux densities at 12 and 25 //m for the stars that were detected both by IRAS and ISOO (Tables 7.3 & 7.4) are compared in Fig. 7.2. A bright regime where ISO and IRAS are consistentt can be distinguished from a faint regime where ISO flux densities are systematically lowerr than IRAS flux densities. CAM is consistent with IRAS down to fainter levels (~ 0.2 Jy) thann PHOT (~ 0.6 Jy). PHOT seems to under-estimate flux densities at levels between 0.2 and 0.66 Jy by a factor ~two. Below 0.2 Jy, both CAM and PHOT yield flux densities ~ 0.6 x IRAS.

Fluxx density under-estimation may be caused by the difficulty of the detectors to respond to loww signals. CAM 12 \x flux densities below ~ 0.2 Jy may have been under-estimated as the K-O methodd does not adequately correct for non-stabilised signals if stabilisation is not reached well withinn less than half the integration time. The 12 and 25 /j,m PHOT measurements, although employingg different detector materials, show exactly the same trend. The 12 (either CAM orr PHOT) and 25 /im observations were performed in the same orbit. Default responsivities


< <

PLH H

1.5 5

uu 1

o o

fcfc 0.5

0 0

" 5 — I — I — I — I — I — I — I — I — I — i — I — I — II I I I I <~

// x

44 <

7t* 7t*

o o

oo o-XX O

6>6> x

ii i i I i I m m ii i i i i

ii 1 1 1 r

= 12 Ai m (CAM)

oo = 12 Ai m (PHOT)

xx = 25 / xm (PHOT)

.. o...

00 0.2 0.4 0.6 0.8 1 55 10

F1So(Jy) )

Figuree 7.2: Comparison between ISO and our revised IRAS flux densities.

2 2

1.5 5

1 1

0 .5 5

0 0

forr the 12 and 25 /im PHOT measurements were thought to be (much) lower in 1996 than inn the current calibrations. Adopting those early values, the ISOPHOT photometry would be consistentt with the IRAS photometry to a high degree. Subsequent revisions of the default responsivitiess have lead to higher values, approaching the values we obtain from the (chopped) FCSS measurements. The ratio of ISO and IRAS flux densities at 25 ^m is 0.60 6 for the 24 sourcess with ISO measurements before orbit 190, and 0.68 0.09 for the 21 sources measured afterr orbit 190. These ratios are very similar, despite the large differences in median IRAS flux densityy between these two samples: 0.41 and 0.14 Jy, respectively.

However,, the discrepancy between the ISO and IRAS data may not be as great as it appears iff we take plausible selection effects into account. The stars in our sample were largely selected onn the basis of their IRAS flux density, but many of them were only just detected by IRAS. Itt is therefore likely that they were near the maximum of their pulsation cycles at the time off the IRAS observation. In contrast, they wil l have been at random phases when the ISO observationss were made. This wil l lead to a systematic difference between the IRAS and ISO fluxflux densities for faint sources. A similar effect may explain the discrepancy between the PHOT andd CAM behaviour for sources with flux densities in the range 0.2 to 0.6 Jy, as the brighter sourcess were selected for measurement with PHOT and the fainter sources with CAM. Ground-basedd 10 yum (N-band) magnitudes of a subset of our ISO targets were on average ~ 30% fainterr than measured by IRAS at 12 u.m (Chapter 3). Although we explained this in terms off differences between the N-band and IRAS 12 /mi filters, it may actually reflect the same discrepancyy seen between the ISO and IRAS photometry. Variability cannot be the complete explanation,, though: for instance, the sources IRAS04407-7000, 4516-6902 and 05003-6712 weree all near the maxima in their K and L-band lightcurves at the time of the ISO photometry, yett their PHOT 25 pm flux densities of 0.584, 0.380 and 0.210 Jy, respectively, are still fainter thann the IRAS flux densities by factors of 0.77, 0.69 and 0.64, respectively. Interestingly, Reid ett al. (1990, their Figure 5) show that for F12,2s^0.3 Jy both the PSC and Schwering & Israel (1990)) over-estimate flux densities for point sources in the LMC by typically 20 to 50%. They attr ibutee this to source confusion resulting from the large beam width and crowdedness in typicall LMC fields.


n n c c

0 0

V V

u u

-̂-

F F

" O O <]> >

3 3

u. .

> > 5! ! < < u u

CO O

h1--O O X X 0--(3 3

-C C

< < u u y y o o

£22 J oo u

gg 5 II v auu — EE g

OSS w

TT u

t33 u >> £ cc c uu u

—;; C ?-.. —

ii I ww 5 £ > CC 00 55 "

oo + a.. .a

E JB JB 5--

a..s s c33 O

inn o

-«I I SS ' « ^QQ t ^^ ti s aa < 8. HH CO as

10 0

ë ë i i

& &

u. .

> > u u s s < < u u

u. .

> > u u 2 2 < < u u

co o

fc fc I I a. .

w w

Is s X X a. . £2 2

Cfll CO CO CO

ëë ë ëë a.. a. a. a.

ëë ë

000 <N 00

2 3 3 3

oo 5 o

i nn * - t o * »rtt I N r^ r^ r^i m mm * - o — — oo o o o o o

uii S — x

r-- * r-~ K> »nn — d m m i n O* m r~ r**j — —— — M — \ 0 O — — O O — —

S R 8 8 e ii — (N —

88 2 5 3 3

oo o o o o o

ooo (N in >o en r̂ § —— — <N — <N

cc O o o o dd o o o o o Pff — 'S St ï ï S oo r~- o m oo

r jj — n o O 66 o d 6 o ' d

mm —

i nn —

00 O O <N O O N - y ^ ï --G** (N Tf v~i O -t 011 o d vO — <o oo o o t~ d o

occ oo — * f , r~ MM N V I i n — (N oo o o o o o dd o o o o d

NN n (N 00 oo — o o dd d d d

r-- r- r- — ( N OO — —

(N N (N N O O

tr, tr,

(N N O O (N N O O

s s

££ i n ooo o rr,rr, —

oo -<r

(NN r-

3\ \ sa a O O O O

<N N i i o o

r ) ) O O

(N N 00 0 o o O O

oo o o c hh - m f 5

d o o d d

oo o o o d o d o o

ooo O 1/1 f̂ -

OO O O f N

— t u ~ i r ^ ( N — — —

888 8 8 8 8 88 d oo d d o ' d o d ; tt m in — oo in fN) S

ess * * M r̂

88 S 8 88 CC ó d o d

oo o o o o o nn o — — cc d o d

(NN 00

88 8 oo o 33 'O (NN 00 dd d

t N m r ^ - * n m t s c s ( N m <N N

8 8 8 8 8 8 8 8 8 8 8 8 d o d d d o o d o o d d SS^ o l ï t o \ Ó c A t £ * h h

« B O - i o - o i f i N ' t i n n

ö ö ö ó ö ö ö ó ö óö ö

OO O O O O ?? o o o o

S mm ( N • * ( N CM — OO O O O O O

dd d d o ' d mm m o m i n o oo — p — o o o o o o o o o —

o d d o d o d c d d d d d d OinOQininO'n'^?'n''n'in O O ooo — —i o — ^ t o o ̂ o O r~ - r - i n i —

o o o o o o r ^ o ^ s o a v o o o o r ^ o ö K r ^ r ^ o d K r ^ o d r ^ o d c ^ o o o d o d j o ' ^ o d o d o d ^ ' '

_ _ o o o o o o o o o O O O O O Q O O i n O O 0 0 £ > r < ;; — r - - * ü r - — (S ï

p pp — p p p r s M p — p p p q q q p — p p p o c o o o o p — p p p d o o — c o o o o o o o d d d d d d d d d o d d d d d d o d d d c c <r<r ii 00 ' O O O o o O C m O ' n ^ m c ' o o i n i ^ O Q p r Ï Q o i n o > O v O o o r ^ r - ; T t 2 F J R o q — o — öb f i iS c» * * — N v i o i i ^ - ( N .

o o o o r - t - -- — — — o o a i o d o d v o K K — — — : o ó o \ a ' o 6 o d c ^ o b o ^ — o\ nn t * h dd r i j o

OO (N C ONN — 55

NN N O m m

oo o — o o OO O (N o oo m m in in in OO O O O

2 2 S S 2 . S 2 £ 2 S 2 S £ ï 9 . 2 0 é ë £ S c c ' i ; ' c ' c ' c ' 0 ' i : 0 d d ^ * 5 2 i n ^ p q r ^ p ^ ^ o o ^ o S o r ^ o o o m o ^ o o ^ S S ^ K S ^ i ^ r ! > > = ' r : ? ? v"ii «n o o o o *

—— O (N (S (N \0 O o ö r - h o ó r i n m o ó d d d d

> n > n > n o o > n O ( r i C C pp — — C — o —

o d d d .. _ _ _ _ > n o ö i n » n ö c i n o c ö

j i d o ó - ' ï - i N ' t d i N T t ^ ' t r i r l n n

mm (N o in >n f i oo o — o o o oo d o d o d oo ro c £5 3 oo

MM O (N <N — rr, — oo — o o o o o oo o d o o o d KK o c^ ^ f o V? r̂ T (NN \0 — r n — v i 3N dd O O O^ 00 00 00

dd m m i n in i n ro in o o o o o o o o o d o o d d o o d d

OO — r̂ t <N O — O d

—— ^ r - ' - ' t T O O i O a v O O t N ^ ^ r < r i —— r ^ o o i n m o o c N r N o o c N i n i n o ^ 5 c*">r~-mm — r^ i n in «o

%oo — ^ o — ^ - s - ^ ^ ^ — f̂ ) — o > — r - < N

—— — O O ( N 3 v O r * l C - l < N 0 0 < N 0 C ( N r M ( N ^^ >£ w w i i V ' vm I t ^ ^ * I N * 1̂ 1 t>l —— ^ N (S (N (N N — v O r ^ C N f N f N ^ O ( N » ü r ^

8 8

-- o —— [^ mm o OO (si

22 è > coo O I

OO Q <7> mm ^ r-(NN d m —— IN (N >> > > X I I I

II I 300 l ~

r-- oo o <o 3 3

I I I I Ï Ï Ï

NN N N O - O ** (N O — M Q ooo 5> c*i a\ oo o

** * * O * h II I I I I I

oo o vOO o o in OO Ö — er, d TJ-tt m m i n i n i n

3 3 3 3 3 3 3 coo co co co co co < < < < < < < XX CC 0£ « OS K

coo oo co << < < ass os os


PL L

COO ( J

ii < o.. U

ë ë 0--

> > s s < <

eoo u t o

00 S O 11 f-, = a.. U cu

u. . > > u u s s < < u u

u--> > u u s s < < u u

Ttt Ov Ov M 9S

-- f- J 2S ° ©© CS O O —

OO O O . o , g . o , 5TT («ï r f OO 9> O r -- > — ! £ ü t -oo es ^ o * —

r -- m —

Ttt — go —— en © —— i n (S

vpp oo OO es öö o

SI I

CSS Ov ooo cs

css m Q

o oo o oo vo 5 —— — CS

22 s

s s

oo o 8 o o ,, © , o © o^^ r-^ i n o '

^ -- — c s —— CS o oo o o

011(

0 0

003(

0 0

00

7(0

0 OO c?< ( S o o

mm © c s — OO O O O

oo o o o. c XX oo cfc f^ —— © cs r--oo o o o

r -- i n T f —

8 S 5 5 5 o,, o o o , r^^ ^5 —̂ in' mm es m © OO O o O

vOO oo ooo oo r -- — oo O O O oo o o o

33 5

88 8 oo o i? ff — css r— nn —

** r ) m

8 8 8 8 oo o o css ï? è? —— O Cl

U i ll 8 8 8 8 8 ©© o d d d o ó d o cJJ èsT - » <<i i ^ » » »

8 S 8 S 8 S 8 8 2 2 ö ö ö ö ö ö ö ö © ©

8 8 8 8 8 oo o, o, o,

©© O © Ö

m m © © © m © o o O O C S C SS — o c s c s £ 3 3 mm in O O * in

OO O — — O O

© © © © m w i o m e s m m SS - - — O O — © O ©

© ö ö © © © © o © © o © o © © , o, , SSw i n è 3 ' o o i K o o o o i n i n S ' oo © O

r n ^ 5 mm — e s r " - r n m © — es o co >£> cs " o \ a ó o ó o « o d o Ö C T i o » o c J r » t ~ - C > Ö r - - o d d

m m o o o m o o i n r n f i m m © © c ss — — © es es o o o o_

© © _ © © © , © , ,

mm o « I 2 n ^ O^^ ON OÖ OO © OÓ

©© © ©^ O , © O , ©© ©^ © © C ? © ^ © ' © ' © .. - . — — - - - i . m i n c s m ^ i n i n i n r n ^ o o © o ó o o o ^ o o o c j o v o v i n r ^ ^

© 0 © © " Ö © © © © © © © 0 © 0©© © © © o o © © . © , © , © , © , © , © ,

g £ S 2 8 s I s 8 g S 8 8 s l ? © :: r ~ ~ ~ ~ - - — > '' — — — CS O — - N M O i o v c i f O a J d

© © © 0 0 © © m m O © ^ eo o

© Ö © © ' © 0 © © © © © _ ©© ©

© ' © o § © ' © o ' ^ © © S o o o ( v i ^ M f i m i / i r ii — — l O v t i Ö —

© © © , , in n r s s O O

in n O O

i n n © © o o 8 8 o o

o o

<n n © © o o © © CS S © ©

©

CS S © © © , ,

oo o © ©

© ©

CS S © © © . . PO O CS S

o o

cs " " © ©

© ©

© ©

— — © ©

© © CS S © ©

8 8 ~ ~ ©CS S

i n n © © © © i n ' '

© ©

i n " " O O

i n n

o o o o in' ' O O

© ©

i n n © ©

o, , rn rn

o o

©

i n n

o o © © i n n

© ©

S' S'

© ©

© © © © CS S a\ \

O O

© ©

© . . © © 90 0 00 0

O O

m m © © © © in n i n n

= = © ©

s s © ©

© ©

© ©

CS S © ©

o o CS S © ©

cs " " © ©

e n n o o

o, , irT T < n n © ©

i n n O O

© © © . . i n n

© ©

S S

CS S © © © , , oó' '

\ C C

CS S

o o o o

oo o

CS--

© © 3 22 _ _ © © © ö © © ' © ö © © o © © © © o © © o , © , o© ©

£ 8 ? i W ^ S r ? S 8 ^ 8 i i S f S g S f S S S « 2 2 ( S r i - -- — — — o i — p i — N O — o i o > ' | i f , i " , ' " " " f k O *

© 0 © m © m c s © © © —— CS — © — © © — CS

22 8 © ' © '' © © © © © Ö Ö © Ö ö © 0 0 _ C ^ O O © © © © , © . 0 . OO CS o O O ^ S m t ^ Q Q O O O O O g Q r J O i c i O i i o O

—— e i ^ ^ c i n n p i — f i c i N n - o i d o i i J v i n f i n s i

J S ö S S S o S o -- v o o o t - c s © o o o o — es m es <s © oo oo . i — S P S S ^ ^ S E S E S S S r n r -- es es r - es — r s r - - c s c s v 5 r - — t s r s r c - > e s r - e s r - f i r - r - c s c s c s r - r r - r - r - t - r ) *o

II I I

o o

o o V) V) < f f fC fC

a> >

o o C/l l

< < U. U.

O O

i n n O O c/ l l <J* * of f

> > c l l i n n © © C/5 5

< < a a

CS S C l l i n n © © CO O

< < Cé é

* *

m m i n n © © CA A

< < f * *

i n n © © Cfl l

< < 0£ £

© ©

o o t r t t

< < a£ £

o o i n c s r ^ O v c s v o m m

o>> q \ ©

mm m — cs —— cs cs

v o v o r ~ r ^ * o r ^ r - r --II I I II I I

3*? ? O > © c i o o i nn — m e s vpp >n cs — iO ~ —

FF o © — i n M - M N ' t ' O u » !! I © c s e s c s c s c s e i r ' " ! © © i n i n m i n i n i n m i n f n n © © o o © o o © _ _ > > > > > > > > P P

i nn cs oo TT Ttt r - oo —

ess es es S

33 o oo So ii x oo o

92 2

N N 10 0

8 8

ChapterChapter 7

ii 1 r " ii r

3a a ii r

6 6

0 0 0 0 0.5 5

F F 600 /um (Jy) ) Figuree 7.3: Histogram of the distribution of ISO 60 (im flux densities (chopped measurements). Thee dotted vertical line indicates 3-cr flux density derived from the distribution of negative flux densities.. Al l flux densities over 1 Jy are piled up in the last bin.

AA histogram of the distribution of ISO 60 /im flux densities (Fig. 7.3), leaving out the mappingg observations, illustrates the detection rate. Considering negative flux densities in-dicatingg non-detection, and assuming a Gaussian distribution around zero flux density for non-detections,, we estimate a 1-cr detection to have 0.103 Jy. There are 12 sources with flux densitiess exceeding 3-cr, i.e. probable detections. This does not take into the account the large errorss on some of the individual measurements, and a 3-cr detection may still turn out to be spuriouss (an example is WBP14). On the other hand, the distribution below 3-cr is certainly skewedd towards positive flux densities. Projecting the negative flux density distribution onto thee positive domain, we estimate that there are probably 17 more detections between 0 and 3-cr,, and a total of 14 non-detections.

Inn the IRAS 60 (im data we found 8 detections and 17 tentative detections (Table 7.3). Thee 0.1 Jy assigned to the faintest IRAS 60 //m flux densities compares well with the ISO \-o detectionn threshold of the chopped measurements. Of the 8 IRAS 60 \xm detections 6 have ISOO chopped measurements, all of which yield higher flux densities than IRAS — by a factor 1.88 on average. This is in contrast to the 12 and 25 /im photometry, where ISO flux densities aree generally lower than those measured by IRAS. The 14 ISO chopped measurements of IRAS tentativee detections also yield higher flux densities than did IRAS — by a factor of 1.5 on average,, although some individual ISO measurements are fainter than the IRAS ones. None off these ISO measurements is negative, indicating that many of the IRAS 60 /v,m tentative detectionss are indeed real.

Thee 7 mapping observations all agree with the chopped measurements within 2-CT, although thesee errors can be large. There is no tendency for one of these two methods to yield higher fluxflux densities than the other. As we do not expect strong variability at 60 /*m, which traces cooll dust some distance from the stars, flux densities from mapping and chopped measurements aree averaged. The error estimates of the mapping measurements are systematically larger than


thosee of the chopped measurements. This may be due to the fact that, for the mapping data, thee flux density of the star was determined from the inner 3 x3 pixels. The contribution of the backgroundd to these 9 pixels is considerable. Also, the reliability of the error estimate for the centrall pixel in the chopped data as produced by PI A is unknown. There is great difficulty in extractingg reliable photometry and associated errors from either mapping or chopped measure-mentss at 60 /xm, for stellar sources in fields like the LMC. This is mainly due to the complex backgroundd and limited spatial resolution of PHOT at these wavelengths. IRAS05289-6617 hass a very smooth background, being situated in the line-of-sight to the supergiant shell LMC4 (Meaburnn 1980). Indeed, ISO mapping and chopped measurements are relatively precise for thiss source, and agree nicely with the 60 /xm flux density measured from the IRAS data.

7.66 Discussion

7.6.11 Chemical types from ISO spectroscopy

Thee presence or absence in the ISO spectra (Figs. 7.4 & 7.5) of discrete dust emission and molecularr absorption bands can be used to distinguish between carbon- and oxygen-rich cir-cumstellarr envelopes (e.g. Merrill & Stein 1976a,b,c). The results are summarised in Table 7.5. .

Amorphouss oxygen-rich dust may give rise to strong and broad silicate emission between ~~ 8 and 13 /xm, peaking at ~ 9.7/xm (the exact location may differ from this by a few tenths off urn). The late-M stars HV2446, 888, 996, and SP77 30-6 have prominent silicate emis-sion.. In optically thick cases the silicate feature turns into absorption. All spectra of OH masermaser sources show the silicate feature in self-absorption: IRAS04545-7000, 05298-6957, 05329-6708,, 05402-6956, and WOH G64.

Oxygen-richh molecules do not provide clear diagnostics of the chemical type of CSEs at our signal-to-noisee and spectral resolution. We already mentioned that shallow absorption around 33 /xm in oxygen-rich sources is most likely due to an artifact in the responsivities, rather than H200 ice.

Crystallinee carbon-rich dust sometimes gives rise to a SiC (graphite) emission feature peak-ingg at ~ 11.3/xm, and narrower than the silicate feature. The CVF spectrum of IRAS05289-6617 (Fig.. 7.4) shows the best example of this.

Carbon-richh molecules have several strong absorption bands in our spectral region, all from HCNN and C2H2. The strongest is at 3.1 /xm, but the problem with the responsivities limits the numberr of unambiguous detections to one (IRAS04496-6958). Related, but weaker, absorption iss visible at 3.8 //in. More absorption bands are located around 5, 8 and 14 /xm. Unfortunately, thee 5 /xm band falls entirely in the blind spectral region of PHOT-S. The 8 and 14 /tm bands aree at the edges of the CVF spectra and hence difficult to identify.

Somee other spectra show merely a featureless dust continuum around 10 /xm. Best examples aree the CVF spectrum of SHV0500193-681706 and the PHOT-S spectrum of IRAS05568-6753. Thesee spectra suggest pure amorphous carbon dust emission.


0.06 6

33 0 .04

& H * 0 . 0 2 2

0 0

0.1 1

0 .05 5

0 0 0.15 5

0.1 1

0.05 5

0.2 2 0.3 3

0.2 2

0.1 1

0 0

0.4 4

0.2 2

0 0 0.1 1

0.08 8 0.06 6 0.04 4 0.02 2

0 0

I-HV12070 0 -z-z HV2379

ii i i i ' i i i i i i

HV2446 6

r IRAS05348-7024 4

11 1 I I I I I L_L _!_ _

IRAS05289-6617 7

OjJ>«-OjJ>«-

JJ 1 I I I L J_ _ JL L IRAS04496-6958 8

o.oo o

J — I — II I I I I I I I

IRAS05298-6957 7

J — i — i — i — ii i i i i i i i i i

IRASO5558-7O00 0

JJ 1 1 1 1 1 I I L_J I I I _L L SHV0500233-682914 4

-i—I—i—i—i—I—i—ii L J i i i_JL3E

II I _L L

IRAS05402-6956 6

J — II I I I L

SHV0500193-681706 6

J—i—I—i—i—ii I i i i 1 i i i I -

;EE SP77 3 0 -6

_L L J — i — ii i i i i i i i i

•• • • •_ w

0.03 3 0.02 2

0.01 1

0 0 0.4 4 0.3 3 0.2 2 0.1 1 0 0

1 1

0.5 5

0 0

1 1

0.5 5

0 0

0.3 3

0.2 2

0.1 1

0 0

0.3 3

0.2 2

0.1 1

88 10 12 14 8 10 12 A(yu.m) )

Figuree 7.4: The CAM-CVF spectra of obscured AGB stars in the LMC. Open symbols represent spectro-photometricc points that are prone to have flux densities that are over-estimated due to stabilisationn difficulties. The spectral shape is best represented by the solid symbols (squares for thee short-, disks for the long-wavelength region). Emission and/or absorption centred at ~ 9.7 fimfim is indicative of oxygen-rich dust (e.g. IRAS05402-6956 and SP77 30-6). whereas carbon-richh dust may show emission at ~ 11.3 /im (e.g. IRAS05289-6617). A featureless continuum aroundd 10 /xm also strongly suggests carbon-rich dust (e.g. SHV0500193-681706).


0.2 2

0.1 1

-0. .

0.5 5

HV2379 9 -- HV888

0.5 5

IRAS04496-6958 8

/ ,, /•"-"

_l_ _ IRAS05003-6712 2

to^nl. to^nl. 1 1 JJ , I L_L

IRAS04545-7000 0

„„ iK W

_L L

WArfTAT T ii i i

IRAS05112-6755 5

>'!,, l' \Aff»I

J J

JJ , I i _ L

0.5 5

0 0 0.5 5

0 0

1 1

0.5 5

- [ — , - - - ] ' '

mmf mmf it* it* ii i i

— - o o

0.5 5

88 10 12 44 6 A(/U.m) )

88 10 12

Figuree 7.5: The PHOT-S spectra of obscured AGB stars (and RSGs) in the LMC. Emissionn and/or absorption centred at ~ 9.7 jLtm suggests oxygen-rich dust (e.g. HV888 andd IRAS04545-7000). Absorption at 3 /urn is seen in carbon star photospheres (e.g. IRAS04496-6958),, but artifacts in the PHOT-S responsivities also mimic weak depression att 3 ^m in the spectra of unambiguous oxygen-rich stars (e.g. IRAS04545-7000).


-- 0.5

Figuree 7.5: (continued).

7.6.22 IR colour-colour diagrams

Thee ISO 12, 25 and 60 fim filters are similar but not identical to the IRAS filters. As the zero-pointss of these ISO filters are unknown, we adopt here the IRAS zero-points. This results inn the following definitions for the (not colour-corrected) mid-IR magnitudes

[12]=-2.51og(F12/28.3) ) (7.1) )

[25]=-2.51og(F25/6.73) )

[60]== -2.51og(FM/1.19)

(7.2) )

(7.3) )

ISOISO observations of AGB stars in IMC 97 7

Tablee 7.5: Chemical types. Optical spectra (Opt Sp) include objective prism and CCD spec-troscopyy up to ~ ljxm. ISO spectroscopy (ISO Sp) comprises PHOT-S and CAM-CVF obser-vations.. IR colour-colour diagrams (IR col) can in some cases be reasonably conclusive too: we herehere use (K - [12]) and ([12] - [25]) versus {K - L) diagrams. At radio wavelengths, OH, SiO and/orr H20 maser emission is detected from some oxygen-rich sources.

Star r GRV0519-6700 0 HV12070 0 HV12501 1 HV2360 0 HV2379 9 HV2446 6 HV5870 0 HV888 8 HV916 6 HV996 6 IRAS04286-6937 7 IRAS04374-6831 1 IRAS04407-7000 0 IRAS04496-6958 8 IRAS04498-6842 2 IRAS04509-6922 2 IRAS04516-6902 2 IRAS04530-6916 6 IRAS04539-6821 1 IRAS04545-7000 0 IRAS04557-6753 3 IRAS05003-6712 2 IRAS05009-6616 6 IRAS05112-6755 5 IRAS05113-6739 9 IRAS05128-6455 5 IRAS05190-6748 8 IRAS05289-6617 7 IRAS05294-7104 4 IRAS05295-7121 1 IRAS05298-6957 7 IRAS05300-6651 1 IRAS05329-6708 8 IRAS05348-7024 4 IRAS05360-6648 8 IRAS05402-6956 6 IRAS05506-7053 3 IRAS05558-7000 0 IRAS05568-6753 3

Optt Sp C C MS3/9 9 Ml.5 5 M2Ia a C C M5e e M4.5/5 5 M4Ia a M3Iab b M4Iab b

C C

M10 0 M9 9

ISOSp p

oxygen? ?

SiC? ? silicate e

silicate e

silicate e

SiC? ?

car+sil? ?

silicate e

silicate? ?

carbon? ?

carbon? ? carbon? ? SiC C

silicate e

silicate e SiC C

silicate e

silicate e carbon? ?

IRcol l carbon? ? ? ?

oxygen n oxygen n carbon n oxygen n oxygen n oxygen n oxygen n oxygen n carbon n carbon n oxygen n carbon n oxygen n oxygen n oxygen n oxygen n carbon n oxygen n carbon n oxygen n carbon n carbon n carbon n carbon n carbon n ? ?

oxygen n carbon n oxygen n carbon n oxygen n carbon n carbon n oxygen n oxygen n oxygen n ? ?

Maser r

yes s

yes s

yes s

yes s

yes s


Tablee 7.5: (continued). Star r SHV0454030--SHV0500193--SHV0500233--SHV0502469--SHV0521050--SHV0522023--SHV0522118--SHV0524565--SHV0526001--SHV0530323--SHV0535442--SP777 30-6 TRM45 5 TRM72 2 TRM88 8 WBP14 4 WOHH G64 WOHH SG374

-675031 1 -681706 6 -682914 4 -692418 8 -690415 5 -701242 2 -702517 7 -694559 9 -701142 2 -702216 6 -702433 3

Optt Sp C C

C C C C M3 3 S? ? MS5 5 C C M6 6 C C M8 8

C C C C

c c M7.5 5 M6 6

ISOSp p

carbon n SiC? ?

silicate e

silicate e

IRcol l carbon n carbon n carbon n carbon? ? carbon n ? ?

carbon n ? ?

carbon n oxygen n ? ?

oxygen n carbon n carbon n carbon n carbon n oxygen n oxygen n

Diagramm of (K - [12]) versus [H - K)

Thee (K - [12]) versus (H - K) colour-colour diagram separates carbon- from oxygen-rich stars inn samples of obscured stars in the MCs (Paper II & Chapter 3). Indeed, the distributions off carbon- and oxygen-rich stars using ISO and SAAO photometry define clear sequences in thiss diagram (Fig. 7.6). The sequences are fit by eye, with the carbon sequence the same as in Chapterr 3:

{H-K)={H-K)= 0.36 x(K- [12]) (7.4)

butt the oxygen sequence a simple, yet somewhat steeper function than in Chapter 3:

(H(H - K) = 0.3 + 0.0003 x (K - [12])5 (7.5)

Althoughh the stars with spectral type M follow the oxygen sequence very well, the carbon starss show a large scatter around the carbon sequence with several carbon stars on or beyond thee region populated by M stars, at small (H - K) but large {K - [12]) magnitudes. This scatterr contrasts with the tight carbon sequence that is observed in the Milky Way (Fig. 3.3). Wee suspect that this is in part caused by the severe crowding in some LMC fields, affecting thee near-IR aperture photometry. Differences in the strength of absorption in the H-band by carbonaceouss molecules may cause additional scatter among carbon stars (Bessell & Wood 1983;; Catchpole & Whitelock 1985).

Diagramm of {K - [12]) versus (K - L)

Thee (K - [12]) versus (K - L) colour-colour diagram shows much less scatter around well-definedd carbon and oxygen sequences (Fig. 7.7). This makes it a much more powerful diagnostic diagramm than the (K - [12]) versus (H - K) diagram in typifying the chemical composition


10 0

8 8

6 6

03 03

W W

nn 1 1 r "ii r ii 1 1 1 1 i i Ti r

G G

•• X

0 0 o o

o.--x x

D D

•• = oo = GG —

AA —

XX =

CC s ta r MM s ta r MSS s ta r SS s ta r unknown n

II I I I I I I I 1 1 L II I I L

0 0 2 2 (H-K ) )

Figuree 7.6: (K - [12]) versus (H - K) diagram. Stars are distinguished by their chemical typess inferred from spectroscopic methods: carbon stars (solid disks), M stars (open disks), MSS stars (open squares), S stars (open triangles), and stars of which the chemical type is a priorii unknown (crosses). Carbon stars and oxygen stars define sequences in this diagram, indicatedd by a dotted and solid curve, respectively.

off the circumstellar dust. Noguchi et al. (1991a) introduced a very similar diagnostic using (L(L - [12]) and (K - L) colours. We note, however, that some of the peculiar stars in our (K(K - [12]) versus (H - K) diagram were too blue and hence too faint to be detected in the L-band.. Still, the tight sequences prove that both the SAAO and ISO photometry are of good qualityy when comparing individual stars. We fit (by eye) a linear carbon sequence:

(K-L)(K-L) = - x (K [12]) ) 2 2

Ï Ï Ï (7.6) )


10 0

CM M

8 8

6 6

0 0

1 1

--

--

--

--

--

--

—— 4 --

V" " 1 1

11 1 1 1 1 1 1 1

o o

'' '

X X

^ ^^ • x •

XX •S-"é' oo • ,•

/ •• .••'

• • 11 1 1 1 1 1 1 1

•• = OO =

DD —

AA =

XX z =

11 1

11 1 1 1 1 1 1

^^ -

~ ~

--

CC s t a r MM s t a r MSS s t a r SS s t a r unknown n

--

--

— —

--

--

ii i i i i i i

00 1 2 3 (K-L) )

Figuree 7.7: (K - [12]) versus (K - L) diagram. Symbols as in Fig. 7.6.

(7.7) )

andd a superposition of even polynomials for the oxygen sequence:

(K(K - L) = 0.35 + 0.007 x (K - [12])2 + 0.0014 x (K - [12])4

Diagramm of ([12] - [25]) versus (K - L)

Anotherr colour-colour diagram that separates carbon- from oxygen-rich sources is the ([12] -[25])) versus (K - L) diagram (Fig. 7.8). The confirmed oxygen-rich sources show a linear relationshipp between the ([12] —[25]) and (K-L) colours, possibly flattening out at (K-L) > 1.5 mag.. The LMC stars generally follow the separation determined for galactic stars (dotted line inn Fig. 7.8, taken from Epchtein et al. 1987).

ISOISO observations of AGB stars in LMC

02 02

CQ CQ

~\~\ r ii 1 1 r 11 1 r

0 0 0 0

•• =

oo =

AA —

XX —

CC s tar MM star SS s tar unknown n

oo o

o o

oo oxygen o o o o

— —

--

1 1

o o o o o

oo ö

o o o o o o

o o 11 1

..©'' '

.. 1

<g» »

, A , ,

o o

carbon n

II I I I I I I I I I I 1 L_

2 2 (K-L ) )

Figuree 7.8: ([12] - [25]) versus {K - L) diagram. Symbols are as in Fig. 7.6. Oxygen-rich sourcess and carbon stars occupy distinct areas in this diagram. The dividing line (dotted) betweenn stars with carbon- and oxygen-rich dust is taken from Epchtein et al. 1987.

Diagramm of ([25] - [60]) versus ([12] - [25])

Thee aim of obtaining 60 u.m flux densities for stars in the LMC is mainly to probe the coolest circumstellarr dust. The 60 /im flux density is expected to increase as prolonged mass loss first extendss the CSE and again as reduced mass loss results in a detached shell. This evolution mightt be seen in ([25] - [60]) versus ([12] - [25]) diagrams (Fig. 7.9, see also van der Veen & Habingg 1988). Unfortunately, the accuracy of the ISO photometry at 60 ^m is not very high forr most of these LMC sources, and the diagram contains a lot of scatter.

Perhapss the most obvious thing to learn from this diagram is that carbon stars tend to


6 6

o o CD D

ii 1 1 r ~ii r ~ii 1—i 1 1 1 1 1 — r " i — i — r r

O O

O O

o o

0 . . JJ L

•• = oo = AA —

xx —

CC star MM star SS star unknown n

o o

o o

o o

•• o o o

o o o o

o o o o o o

o o

JJ I L

o o 0.5 5 11 1.5 ([12]-[25]) )

JJ L

Figuree 7.9: ([25] - [60]) versus ([12] - [25]) diagram. Symbols are as in Fig. 7.6. Carbon stars aree not well separated from oxygen-rich sources, although carbon stars seem to be relatively brightt at 60 fim.

bee relatively bright at 60 /an, yielding ([25] - [60]) ~ 1.5 to 3 mag. Although oxygen-rich sourcess can have similar colours, there are many oxygen-rich sources with ([25] - [60] j < 2 and ([12]] - [25]) > 0.6 mag, colours not seen for any carbon star in our sample. This is similar to thee findings of van der Veen & Habing (1988), but our LM C sources have bluer ([12] - [25]) and redderr ([25] - [60]) colours than do their Milk y Way sources. However, the LMC ([12] - [25]) colourss do not differ much from those discussed by Le Bertre et al. (1994).


7.6.33 Comments on particular objects

GRV0519-6700 0

Thee referee Dr. Peter Wood conveys that an optical spectrum of GRV0519-6700 shows it to bee a carbon star, in good agreement with its IR colours of {H - K) = 0.7 and (K - [12]) = 1.0 mag. .

HV12070 0

Thee CVF spectrum of HV12070 shows only a hint of the silicate feature, whilst the IR colours cannott distinguish between oxygen- and carbon-rich dust of the optically thin CSE of this MS-typee star.

HV2379 9

Thee PHOT-S spectrum of HV2379 suggests SiC emission, but its CVF spectrum does not. This mayy be a result of changes in the properties of the CSE or the dust. Its IR colours leave no doubtt about the carbon-rich nature of the dust.

HV2446,, 888, 996, and SP77 30 -6

Thesee late-M stars all have prominent silicate emission and IR colours that unambiguously

indicatee oxygen-rich dust.

IRAS04286-6937,, 04539-6821, 04557-6753, 05009-6616, 05113-6739, 05295-7121,, 05300-6651, 05360-6648, and TRM4 5 and 72

Thee position of these objects in the (K - [12]) versus (H - K) or {K - L) colour-colour diagramss does not clarify the chemical composition of their CSEs. The ([12] - [25]) versus (K(K - L) diagram, however, unambiguously indicates that the dust around these stars is carbon rich.. The IR colours of IRAS05113-6739 at the three ISO epochs for this star all lie along thee carbon sequences in the (K - [12]) versus {H - K) and {K - L) diagrams. Ground-based L-bandd spectra of IRAS05009-6616 and 05300-6651 show the 3.1 fun absorption feature due too HCN and C2H2 molecules, indicating carbon-rich photospheres (Chapter 4).

IRAS04374-6831 1

Thee position of IRAS04374-6831 in the ([12] - [25]) versus (K - L) diagram indicates carbon-richh dust. Its PHOT-S spectrum, which does not clearly reveal the chemical composition of the dustt by itself, is then marginally consistent with SiC emission.

IRAS04496-6958 8

IRAS04496-69588 shows strong absorption by carbonaceous molecules at 3.1 /mi, already known fromm ground-based L-band spectroscopy (Chapter 4). Related, but weaker, absorption is visible att 3.8 urn, and possibly around 8 /xm. Surprisingly, this carbon star has silicate emission too, indicatingg the presence of oxygen-rich dust (see Chapter 8). Its IR colours indicate carbon-rich dust,, hence the oxygen-rich dust is only a minor component.


IRAS04530-6916 6

Withh (K - L) = 2.13, (K - [12]) = 6.9 and ([12] - [25]) = 2.3 mag, the IR colours of IRAS04530-69166 imply that the dust around this very luminous and red object must be oxygen rich. .

IRAS04545-7000,, 05298-6957, 05329-6708, 05402-6956, and WOH G64

Thesee OH maser sources all show the silicate feature in self-absorption, and also their IR colours clearlyy indicate oxygen-rich dust.

IRAS05003-6712 2

Thee IR colours of IRAS05003-6712 unambiguously classify the dust as oxygen rich. The PHOT-SS spectrum shows a hint of the silicate feature. A ground-based L-band spectrum of thiss star shows a featureless continuum around 3.1 /xm, indicating an oxygen-rich photosphere (Chapterr 4).

IRAS05112-6755 5

Thee dust around IRAS05112-6755 is classified as carbon rich on the basis of the position in thee ([12] - [25]) versus (K - L) diagram. There is a hint of 8 /an absorption in the PHOT-SS spectrum of IRAS05112-6755. A ground-based L-band spectrum of this object shows the strongg absorption at 3.1 fxm found in carbon-rich stellar photospheres (Chapter 4).

IRAS05128-64555 and 05190-6748

Thee absence of clear indications for the presence of the silicate feature in the PHOT-S spectra off these stars suggest that their dust may be carbon rich, which is also indicated bv their ([12]] - [25]) and (K - L) colours.

IRAS05289-6617 7

Thee CVF spectrum of IRAS05289-6617 shows prominent SiC emission. Hence it is probably aa mass-losing carbon-rich AGB star in the LMC rather than a foreground object. We have not yetyet identified its near-IR counterpart.

IRAS05348-7024 4

Thee CVF spectrum of IRAS05348-7024 shows weak SiC emission. The carbon-rich nature of thee dust around this object is also indicated by its position in the ([12] - [25]) versus (K - L) diagram. .

IRAS05506-7053 3

IRAS05506-70533 is the only star in our sample that could not be detected at 12 /zm. Assuming aa 12 fxm flux density < 0.03 Jy, the (K - [12]) colour would be < 6.2 mag and probably ([12]] - [25]) > 1.5 mag. At (K - L) = 3.3 mag, this suggests an oxygen-rich CSE.


IRAS05558-7000 0

Thee CVF spectrum of IRAS05558-7000 is similar to the CVF spectra of IRAS05298-6957 andd 05402-6956, showing silicate emission that is becoming optically thick at 10 /an. The IR colourss of IRAS05558-7000 unambiguously imply that the dust is oxygen-rich.

IRAS05568-6753 3

Thee PHOT-S spectrum of IRAS05568-6753 shows a featureless dust continuum around 10 j/m, suggestingg pure amorphous carbon dust emission. The near-IR counterpart of this object has yett to be identified.

SHV0500193-681706 6

Thee CVF spectrum of SHV0500193-681706 shows a featureless dust continuum around 10 /jm,, suggesting pure amorphous carbon dust emission. The carbon-rich nature of the dust is confirmedd by the position in the (K - [12]) versus (K - L) diagram. Inaccuracy of its 25 nm fluxflux density causes the rather odd position among the oxygen-rich stars in the ([12] - [25]) versuss (K - L) diagram.

SHV0500233-682914 4

Thee CVF spectrum of SHV0500233-682914 shows a hint of SiC emission, and also its IR colourss clearly indicate that the dust around this star is carbon rich.

SHV0502469-692418,, 0522023-701242 and 0524565-694559

Thee carbon star SHV0502469-692418, the M-type star SHV0522023-701242 and the MS-type starr SHV0524565-694559 are surrounded by an optically thin CSE and hence it is difficult to classifyy the chemical type of their dust from IR colour-colour diagrams.

SHV0522118-702517 7

SHV0522118-7025177 was tentatively classified an S-type star by Hughes & Wood (1990). Its IRR colours are clearly similar to those of carbon stars. This suggests that carbon-rich dust dominatess the absorption and emission characteristics of the CSE despite the under-abundance off carbon atoms in its photosphere. Noguchi et al. (1991b) show that the IR colours of the CSE indicatee oxygen-rich dust in case of an MS-type star. Also, CS stars show 3 /zm absorption from HCNN and C2H2 molecules, whereas SC stars do not (Catchpole & Whitelock 1985; Noguchi k Akibaa 1986). This suggests that carbon chemistry is dominant in CS stars, but not in SC stars. Thus,, we identify SHV0522118-702517 with a CS star. Dust-enshrouded S stars — including MSS and CS stars — that have (K - L) > 1 mag are very rare in the Milky Way, and none aree known with (K - L) > 2 mag (Noguchi et al. 1991b). Hence, with (K - L) = 1.3 mag, SHV0522118-7025177 is among the most obscured S stars known.

SHV0530323-702216 6

Thee late-M type star SHV0535442- 702433 has ([12] - [25]) = 1.56 mag. Its near-IR colours aree rather blue and {K - L) is not expected to be larger than unity. Hence the position of this objectt in the ([12] - [25]) versus (K - L) diagram suggests that the dust is oxygen rich.


SHH V0535442 - 702433

Thee carbon star SHV0535442-702433 is surrounded by an optically thin CSE, and hence the IRR colours are difficult to use for classifying the chemical type of the dust. The location among oxygen-richh stars in the ([12] -[25]) versus (K-L) diagram is caused entirely by the inaccuracy off its 25 ^m flux density yielding a spuriously red ([12] - [25]) ~ 3 mag.

7.77 Conclusions

ISOO spectroscopy is used to determine the chemical type of the dust around obscured cool evolvedd stars in the LMC. ISO photometry at 12, 25 and 60 /jm is presented, together with quasi-simultaneouss near-IR photometry from the ground (SAAO). The accuracy and sensitivity off the ISOPHOT photometry is not much better than can be achieved from properly treated IRASS data. The ISOCAM photometry is much more reliable because it is based on imaging, and ann order of magnitude more sensitive than was IRAS. Colour-colour diagrams prove that relative photometryy is reliable. A combination of (K - [12]) and ([12] - [25]) versus (K - L) diagrams providee a reliable way of distinguishing between carbon- and oxygen-rich dust, provided the CSEE has sufficient optical depth. The combination of ISO spectra and photometry enabled us too securely classify the chemical type of the dust around nearly all stars in our sample. This was previouslyy known for only a minority of the stars. Surprisingly, the (K - [12]) versus {H - K) diagnosticc diagram contains a lot of scatter especially among carbon stars.

Manyy of the obscured AGB stars in our sample are carbon stars: 46% amongst the LMC starss that were detected by IRAS (Table 7.1). M stars were always found to be surrounded byy oxygen-rich dust. In particular, all detected OH maser sources show self-absorbed silicate emission.. As in the Milky Way, the fact that no M star with carbon-rich dust has ever been foundd suggests that HBB cannot efficiently turn carbon stars back into oxygen-rich stars. The dustt around the dust-enshrouded S star SHV0522118-702517 has the characteristics of carbon-richh material, suggesting it is actually a CS star. Surprisingly, the dust around the carbon star IRAS04496-69588 has a (minor) oxygen-rich component (Chapter 8).

uva-dare (digital academic repository) mass loss and ......van loon, j. t. (1999). mass loss and...

Documents